最新资讯
stackoverflow.blog 欢迎来到芯片时代:半导体制造、AI 数据中心与芯片设计的未来 www.youtube.com AI Agent 总犯同样的错:CentralAgent AI 将记忆重构为经验系统 pub.towardsai.net Netflix 科学家建议重采样 LLM 评判:5 次仅将误差棒缩小 13% pub.towardsai.net Jev 作为 LLM 前的静默层:构建提示感知的模型路由器 reddit.com OpenAI 称 Astra 仅满足 10% 安全任务标准 dev.to FreeToken:在自有 GPU 上运行前沿 MoE 模型及其内存需求 dev.to 当 AI Agent 说“完成”时,究竟完成了什么? dev.to 部署 Underdog Bark 3.0 Pro:逐字转写、自定义后端与分块设置 dev.to 为艰难对话构建 AI:五个 Agent、一条安全正则与一位盲评监督者 techcrunch.com Palo Alto Networks 洽谈以 5 亿美元收购 AI Agent 安全初创公司 Clutch theinformation.com 黄仁勋被要求就涉华间谍指控向国会作简报 www.infoq.com Edera 推出 Object Capability Bounded Authority,试图将凭证与 AI Agent 工作负载隔离
多元动态
登录 注册
智能体 aibusiness.com · @Esther Shittu

Microsoft Copilot 改版:从助手转向内容工作区

原文 我的收藏

据 AI Business 报道,Microsoft 对 Copilot 进行改版,将其从简单的 AI 助手转变为内容工作区,并为知识工作者提供更高级的开发工具。摘要未披露具体功能细节、发布时间或可验证的实测结果。

AI推荐理由 摘要仅说明 Copilot 从助手改为内容工作区并面向知识工作者提供开发工具,未给出具体功能、版本号或可验证证据,属媒体转述;对关注 AI 应用与办公工具演进者有参考价值,但信息密度有限。

62

正文

Generative AI Data management News Microsoft’s Copilot redesign provides an experimental opportunity The revamp transforms Copilot from a simple AI assistant into a content workspace, providing knowledge workers with advanced developer tools. Esther Shittu,News Writer, AI Business September 28, 2026 4 Min Read Getty Images Microsoft's redesign of Copilot shows how vendors are still experimenting to figure out how to integrate AI with applications that knowledge workers are using. Microsoft reintroduced Copilot with three major new capabilities: Home, Code and Autopilot. Home provides both Teams chat and Copilot Cowork together, with the full Word, Excel and PowerPoint apps integrated into the platform, Microsoft said. Code is powered by GitHub Copilot and lets users build custom apps, dashboards and workloads using natural language. Autopilot is a digital co-worker formerly known as Scout. It can set up and manage review processes, build schedules, prep meetings and follow-ups. The revamp of Copilot, unveiled on Sept. 25, shows how Microsoft is transforming the platform away from being just an AI assistant into a content workspace. With the change, Microsoft hopes to change how customers use Copilot. According to reports, fewer than 20 million users pay for Copilot out of more than 450 million Microsoft 365 customers. Moreover, Microsoft lacked clear branding for what it wanted Copilot to be, and many enterprises were unsure what they were supposed to use it for. Related:AI safety is falling behind the pace of development With the rebrand of Copilot, Microsoft is addressing some of these issues by integrating the Copilot brand with Office and other productivity apps that enterprises use., Microsoft also appears to be trying to capitalize on the popular AI coding domain, much like Anthropic, OpenAI and SpaceX (with the acquisition of Cursor) have done over the past year, while also trying to make Copilot central to how enterprise employees perform their day-to-day jobs. Vibe writing “This is Microsoft now bringing in coding agents, AI coding agents right into a knowledge worker’s experience, which is kind of bizarre or will be an unusual thing for CIOs to see,” said Larry Cannell, an analyst at Gartner. “The coding tab and giving access to an AI coding agent is an industry experiment.” While access to code could work for knowledge workers, what really matters is what is done with the code created, Cannell said. He said that is why Microsoft Copilot Managed Runtime will be key for users, Cannell added. This new platform is a hosting infrastructure that enables code to run safely within an enterprise’s Microsoft 365 environment, according to the tech giant. While IT governs it, Microsoft said it will be easy for users to share an app with teammates, connect it to live data and access it from anywhere. “It’s not creating the code, it’s maintaining the code and supporting it and keeping it running long term,” Cannell said. Related:Gemini 3.8 text-to-speech refines voice AI capabilities With the runtime, though, Microsoft has an advantage because there is an identifier that helps manage the privacy and permissions settings, Cannell continued. “It fits into their overall strategy,” he said, referring to Microsoft’s push to integrate the AI tools within regular applications while still allowing each user to work within the right permission settings. “It is still out there ahead of what most knowledge workers are ready for, so that is why it is an industry experiment. But it is one that Microsoft can compete in well.” Another part of the redesign is that Copilot becomes more central to enterprise knowledge workers. Previously, many users looked to Microsoft 365 as their main application, with Copilot on the side. However, Copilot is now the main application, with Word and other Office apps on the side. “You’re almost going to be doing … vibe writing with this,” Cannell said, referring to mostly natural language-based AI coding. “But you are also going to be vibe co-editing with this as well,” he said. Those types of integrations are going to help the users.” “The real value for most companies that have a Copilot license today is probably going to be the Office integration,” he continued. Related:Anthropic, OpenAI launches show shift toward multi-model enterprise AI Changes in pricing Microsoft is also making a pricing change with the redesign: features in Copilot are billed based on usage rather than at a flat rate. Microsoft said that with usage-based billing (UBB), users have full control over choosing models and matching cost to value. Cowork, Code, Autowork and frontier models such as Astra from OpenAI and Fable from Anthropic all run on UBB. However, what makes Microsoft competitive in pricing is that the vendor is maintaining its tiered licensing model consisting of free Copilot; a $20 per user per month option for Copilot Pro; $21 per user per month plan for Microsoft 365 Copilot Business; and a $30 per user per month deal for Microsoft 365 Copilot Enterprise. “That's good news that has come out here is that they have not gone entirely to consumption. Cost,” Cannell said. “I don't expect them to do that anytime soon.” Esther Shittu is an Informa TechTarget news writer and podcast host covering AI technology. About the Author Esther Shittu News Writer, AI Business Esther Shittu has covered AI technologies and industry trends since 2021. As co-host of the Targeting AI podcast, she talks with experts, thought leaders and practitioners exploring critical AI developments. Before AI Business, she wrote for SearchEnterpriseAI, the New York Daily News, Bklyner and the Brooklyn Daily Eagle. When she's not diving deep into the world of AI, she spends her time on passion projects and raising her three daughters. See more from Esther Shittu Want more AI Business stories in your Google search results? Add Us Now Sign Up for the Newsletter The most up-to-date AI news and insights delivered right to your inbox! Stay Updated!

来源报道 · 1 篇 · 以下链接均前往原始媒体

← 返回精选动态 原文

更多内容